Invasive mold infections (IMI) are a leading infectious cause of death in immunocompromised patients. IMI is most frequently diagnosed in the lungs of children with neutropenia or graft versus host disease (GVHD). Initial clinical suspicion for a pulmonary IMI (PIMI) is often based on lesions identified by chest CT, categorized as ?possible PIMI?. Additional diagnostic testing is usually limited to invasive procedures with concerning risk profiles. The objective of this study is to establish a comprehensive non-invasive diagnostic approach in a prospective multi-center cohort of 300 children with prolonged neutropenia or GVHD that present with possible PIMI. Establishing an accurate non-invasive diagnostic strategy could reduce morbidity from invasive procedures and reduce time to appropriate antifungal therapy, resulting in a reduction in IMI mortality. To accomplish this, we will leverage the International Pediatric Fungal Network, a unique multidisciplinary group of 55 worldwide sites and the only such group dedicated to pediatric invasive fungal disease. This study will benefit from an unparalleled collaboration with the Children's Oncology Group (COG), the world's largest organization devoted to pediatric cancer clinical trials. In Aim 1, we will investigate a comprehensive non-invasive diagnostic testing approach to confirm the presence or absence of proven or probable PIMI in children with newly identified possible PIMI. We hypothesize that a non-invasive diagnostic strategy using available blood-based assays (galactomannan, Aspergillus PCR, and Mucorales PCR) will provide positive and negative post-test probabilities that are clinically informative. In Aim 2, we will leverage this cohort to compare the outcomes for children with possible PIMI managed with empiric antifungal therapy vs. an invasive diagnostic procedure followed by management based on results. Comparison will be using an outcome score measure that incorporates both negative and positive consequences of each exposure, providing a balanced measure of the impact of a clinical decision. We established a minimum clinically important difference (MCID) to dichotomize outcome into `success' or `failure' from surveying experts in clinical mycology. We hypothesize that patients receiving empiric antifungals will have more successful outcomes compared to patients undergoing an invasive diagnostic procedure. This cohort also offers the possibility to study novel non-invasive diagnostic tests not widely available. Therefore, in Aim 3, in a subset of patients we will explore the host response to proven/probable PIMI using RNA-Seq to define transcriptional signatures, and assess plasma cell-free DNA/RNA next-generation sequencing to detect mold pathogens. At select sites, we will also use breath testing to assess volatile metabolite signatures as a marker for IMI. In summary, we will establish the utility of a new comprehensive non-invasive diagnostic approach for possible PIMI using a panel of assays in a unique multi-center prospective pediatric cohort. We will also compare, for the first time, the outcomes of children managed empirically vs. undergoing invasive procedures. These results will provide the foundation for new evidence-based pediatric guidelines.